PhD student, Purdue University
1 paper at NeurIPS 2025
We develop a general theory of agnostic online learning from continuous-time data streams under limited queries, providing tight regret bounds for both oblivious and adaptive settings.
